O
pen
A
rchive
T
OULOUSE
A
rchive
O
uverte (
OATAO
)
OATAO is an open access repository that collects the work of Toulouse researchers and
makes it freely available over the web where possible.
This is an author-deposited version published in :
http://oatao.univ-toulouse.fr/
Eprints ID : 4971
To link to this article : DOI :10.1016/j.jeurceramsoc.2011.01.004
URL :http://dx.doi.org/10.1016/j.jeurceramsoc.2011.01.004
To cite this version :
El Horr, N. and Valdez-Nava, Zarel and Tenailleau,
Christophe and Guillemet-Fritsch, Sophie ( 2011) Microstructure of
Ba1−xLaxTiO3−δ ceramics sintered by spark plasma
sintering. Journal of the European Ceramic Society, vol. 31 (n° 6).
pp. 1087-1096. ISSN 0955-2219
Any correspondance concerning this service should be sent to the repository
administrator: [email protected]
.Microstructure of Ba
1
−x
La
x
TiO
3
−δ
ceramics sintered by Spark Plasma
Sintering
N. El Horr
a,∗, Z. Valdez-Nava
b,c, C. Tenailleau
a, S. Guillemet-Fritsch
a aInstitut Carnot CIRIMAT, Université Paul Sabatier, UMR CNRS 5085, 118 route de Narbonne, 31062 Toulouse Cedex 9, FrancebUniversité de Toulouse, UPS, INPT, LAPLACE (Laboratoire Plasma et Conversion d’Energie), 118 route de Narbonne, F-31062 Toulouse Cedex 9, France cCNRS, LAPLACE, F-31062 Toulouse, France
Abstract
Nano-sized Ba1−xLaxTiO3(0.00≤ x ≤ 0.14) powders were prepared by a coprecipitation method and calcined at 850◦C in air. The corresponding
ceramics were obtained by Spark Plasma Sintering (SPS) at 1050◦C. These ceramics are oxygen deficient and are marked as Ba1−xLaxTiO3−δ. Both
powders and ceramics were characterized by X-ray diffraction (XRD) and Transmission Electron Microscopy (TEM). The effect of lanthanum concentration on the densification behavior, on the structure and the microstructure of the oxides was investigated. Average grain sizes are comprised between 54 (3) nm and 27 (2) nm for powders, and 330 (11) nm and 36 (1) nm for ceramics according to the La-doping level. Powders crystallize in the cubic (or pseudo-cubic) perovskite phase. The structure of ceramics consists in a mixture of cubic (or pseudo-cubic) and tetragonal perovskite type phases. As the lanthanum content increases, the tetragonality of the samples decreases, as well as the grain size.
Keywords: BaTiO3; SPS; Microstructure; Grain size; Electron microscopy
1. Introduction
Since the 1940s, barium titanate BaTiO3 (BTO), pure or doped with different elements has been the subject of many investigations, essentially because of its interesting proper-ties (ferroelectric, dielectric, optical, pyroelectric, piezoelectric, etc.) which make it a very attractive material for the electronic industry1–4(thermistors, multilayer ceramic capacitors, pressure sensors, etc.). As miniaturization drives a branch of this industry, some barriers appear in conventional sintering processes, such as the grain size (GS), the sintering temperature and the purity. Therefore, new routes must be explored to produce this ceramic material.
It is well known that the GS of the material (powders and ceramics) represents one of the parameters which has a large effect on material properties, and has in consequence, a high impact on their specific applications. Several new powder syn-thesis methods and sintering processes have been reported in
∗Corresponding author. Tel.: +33 5 62 25 82 53; fax: +33 5 62 25 82 77.
E-mail address:[email protected](N. El Horr).
the literature in order to control the chemical and structural characteristics and thus the properties of the final product.
Recently, there has been a growing interest for Spark Plasma Sintering (SPS) technique. It offers, in comparison with the conventional sintering methods, the possibility of a very fast densification (generally several minutes only) which limits the grain growth. In other words, SPS is a rapid way for producing dense ceramics with small GS. This method has already been applied to obtain dense ceramics of BTO.5–15It has been shown that the use of SPS technique with fine BTO powders (from several nanometers to a few micrometers) prepared by various methods (hydrothermal, hydrolytic, sol–crystal methods, etc.), and also, the controlling of sintering conditions, allowed us to produce dense ceramics with GS ranging from 15 nm to several micrometers. Also, the SPS technique presents the advantage of using much lower sintering temperature, much shorter sinter-ing time thus leadsinter-ing to inhibition of exaggerated grain growth when compared with conventional sintering methods. The prop-erties of BTO can also be modulated by doping the material with rare-earth elements such as lanthanum (La).16–19If the behavior of La-doped BTO ceramics obtained using conventional sinter-ing is well known,20–28 understanding the effect of La-doping
Fig. 1. Densification curves of Ba1−xLaxTiO3samples with x = 0, 0.01, 0.05, 0.08 and 0.14. The top-left inset shows an enlarged view of curves at the beginning of
the densification. The curves shown in the right part of the figure correspond to time-dependent changes in density for each sample during the dwell time. on SPS sintered BTO ceramics is still in progress.29This work
presents a systematic study concerning the influence of La con-centration on the GS, the structure and the densification behavior of Ba1−xLaxTiO3−δ(x = 0, 0.01, 0.05, 0.08, 0.14) oxides sintered by SPS.
2. Experimental procedure
2.1. Sample preparation
The La-doped BTO oxalate powders were first obtained using a coprecipitation reaction. The calcination treatment led to the oxide powders. The starting materials were BaCl2·2H2O (Pro-labo), TiCl3(Prolabo, d = 1.20, % min = 15) and LaCl3·7H2O
Fig. 2. XRD patterns of Ba1−xLaxTiO3powders with x = 0, 0.01, 0.05, 0.08
and 0.14. The top-right inset represents an enlarged view of XRD patterns for 50◦< 2θ < 52◦showing the (2 1 0) peak displacements with the variation of La content.
(Prolabo, 99%). They were dissolved in water in various pro-portions and the coprecipitation was performed by addition of a solution of oxalic acid dissolved in ethanol. TiCl3was chosen because it was easier to handle than TiCl4. The full oxidation of Ti3+in Ti4+in the solution was ensured by air bubbling during the reaction which was evidenced by the color variation from brown to yellowish. The volume of water was taken much lower than the one of ethanol, in order to decrease the dielectric per-mittivity of the precipitation media. Since the nucleation of the particles is favored compared to their growth, smaller particles, homogeneous in size and composition, are likely to be obtained. The solution was aged for a couple of hours, and the obtained precipitate was centrifuged. The precursors were then pyrolized in air at 850◦C for 4 h to obtain the oxides.
Samples were sintered in a SPS apparatus (Sumitomo Coal Company – Dr Sinter 2080). The powders were placed in graphite die with a lining of graphite paper and heated under vacuum to a sintering temperature of 1050◦C with a holding
Table 1
Density and densification of La1−xBaxTiO3−δceramics.
La content (x) 0.00 0.01 0.05 0.08 0.14 Theoretical density of ceramics dth (g/cm3) 6.0 6.0 6.0 6.1 6.1 Green density d1(g/cm3) 3.0 2.9 2.8 2.8 2.8 Density of ceramics after sintering d2 (g/cm3) 5.8 5.8 5.7 5.2 5.2 Densification (%) 96.7 96.7 95.0 85.2 85.2
Fig. 3. XRD patterns of Ba1−xLaxTiO3−δceramics with x = 0, 0.01, 0.05, 0.08 and 0.14, sintered at 1050◦C, for 44.25◦< 2θ < 46.25◦showing the splitting of (0 0 2) and (2 0 0) peaks and showing the decrease of the tetragonality (i.e., the decrease of the magnitude of the peak splitting) with increasing La content. time of 3 min. The shrinkage of the sample as a function of time was recorded during the overall process. A 50 MPa pressure was applied at 950◦C just after the sintering onset was observed. The pressure was maintained constant till the end of the dwell time at 1050◦C. After that, the pressure was removed, and the sample
Fig. 4. Bright field TEM image of an undoped BTO powder sample showing grains which are in majority equiaxed (Daveragevalue = 54 nm). Elongated grain
is marked by an arrow.
Fig. 5. Bright field TEM images of Ba1−xLaxTiO3powder samples (x = 0.01)
showing: (a) spherical shaped grains, as the grain marked by an arrow, and with GS value∼27 nm; (b) tetragonal (close to quasi-cubic) and cubic shaped grains, as the grain marked by an arrow, and with GS much higher than those shown in (a).
was cooled to the room temperature by shutting down the power supply. The obtained samples were polished with SiC paper to remove carbon deposited on surface samples from the graphite paper during sintering.
2.2. Sample characterization
A Brucker D4 powder diffractometer was used to determine the X-ray diffraction (XRD) patterns for the oxide powders and the sintered ceramics. The diffractometer operated with an emitting source of Cu (K␣1,␣2mean = 1.5418 ˚A).
The sample preparation of SPS sintered ceramics for Trans-mission Electron Microscopy (TEM) observations requires a lot of skills and deserves to be detailed. In fact the SPS samples are very brittle and have a mechanical behavior close to the one of glass.
A first block of sample was cut from the sintered pellets using a diamond saw (ESCIL 3032-4). This block was placed inside a brass tube (diameter = 3 mm) and bound with an epoxy resin (Gatan G1). Following epoxy polymerization (50◦C overnight) the tube was sliced into 500 micron thick discs using the same diamond saw. Mechanical polishing was done up to 100 microns
Fig. 6. Bright field TEM images of Ba1−xLaxTiO3 powder samples: (a) for
x = 0.05 showing tetragonal (close to quasi-cubic) shaped grains for which one or more angles are round shape as the grain marked by an arrow; (b) for x = 0.14 showing that grains are equiaxed and their Daverage(about 28 nm) is lower than
that of x = 0.05 (about 42 nm).
(ESCIL: ESC 300 GTL). At last, a concave dimple polishing (a EA-Fishione – model 2000 – polishing liquid: solution with dia-mond in suspension) followed by ion beam thinning (GATAN PIPS) were utilized to obtain a sample thickness of approxi-mately 100 nm.
The samples (powders and ceramics) were observed using a JEOL JEM 2010 electron microscope operating at 200 kV. The electron beam was emitted from a LaB6single crystal cathode. The sample holder can be tilted around the x-axis or the y-axis and the amount of the tilt is±30◦. The maximum resolution of this microscope is: 0.23 nm point point and 0.14 nm line line.
For each sample, the granulometric analyses of TEM images were realized using ImageJ software.30
3. Results
3.1. Densification
The green density and the densification after sintering of the various ceramics are reported inTable 1. The densification depends on the lanthanum content (x). For a given sintering temperature, i.e., 1050◦C, the highest value of relative density is 96.7% for pure BTO. The densification, then, decreases down to a value of 85.2% for x = 0.08 and x = 0.14.
The sintering curves are shown inFig. 1for each composi-tion. The shrinkage onset temperature also depends on the La content (x): the highest the value of x, the lowest the temperature of sintering onset. This temperature is 762◦C for x = 0.14 and increases to 812◦C for x = 0.01 and reaches 855◦C for x = 0. The kinetic of sintering, that is directly proportional to the slope of the curve, also strongly depends on the composition. In this case, a higher La content leads to a slower densification rate. So, both the onset temperature and the kinetic of sintering explain the decrease of density observed as the La content increases.
In previous work29we noticed that SPS sintering can cause an oxygen deficiency and minimal carbon pollution due to vacuum and the graphite matrix used to press and sinter the powders: CO2desorption experiments show that carbon contamination of the samples is minimal and final ceramics are oxygen deficient (Ba1−xLaxTiO3−δ) when oxygen is dosed by EPMA (Electron Probe Microanalysis) and EELS (Electron Energy Loss Spec-troscopy).
3.2. Structure 3.2.1. Powders
The X-ray diffraction (XRD) patterns of the nano-sized pow-ders are reported inFig. 2. The structure of each initial powder is a cubic (or pseudo-cubic) perovskite. This observation is consis-tent with the fact that BTO crystallizes in the cubic phase, when GSs are fine (belonging to the nanometric scale).28,31–33The pat-terns of Ba1−xLaxTiO3fit with that of cubic BTO (JCPDS file 01-079-2263) pattern with slight shifts in peak positions towards higher angles. They indicate that the lattice parameter slightly decreases with an increase in La content. They are due to the difference in ionic radius between lanthanum and barium (La3+ ion size = 1.36 ˚A and Ba2+ion size = 1.61 ˚A34in 12 coordinate for cubic phase).
The peaks are shifted on the right hand side for 0≤ x ≤ 0.05 and for 0.08≤ x ≤ 0.14. But the peaks are shifted towards lower values of 2θ for 0.05 ≤ x ≤ 0.08; this unexpected behavior could be related to a particular phenomenon reported by different
Fig. 7. Histograms of La-doped BTO samples (powders and ceramics obtained at 1050◦C) showing the change in sample GS distributions (x = 0, 0.01, 0.05, 0.08 and 0.14).
authors.35–39These authors showed that below a critical value of GS (∼80 nm), the structure of BTO is cubic, and the lattice parameter increases as GS decreases and this phenomenon is amplified when GS is lower than 30 nm. So, in our case, the pow-der of composition x = 0.05 has a GS of 42 nm which is relatively high compared with that obtained for x = 0.08 (GS = 27 nm) and the associated increase in lattice parameter is large. Further doping, from x = 0.08 to x = 0.14, did not impact the average GS (Daverage). So, the peak positions were influenced only by the increase of La-doping and were displaced in the expected direction, towards higher 2θ values.
3.2.2. Ceramics
For each ceramic sample, the part of XRD pattern that cor-responds to 44.25◦≤ 2θ ≤ 46.25◦, is shown inFig. 3. For each composition, the splitting of the (2 0 0) peak (2θ = 45.234◦in confirmity with JCPDS file 01-079-2263) into (0 0 2) and (2 0 0) peaks (with the respective values 2θ = 45.092◦ and 45.510◦ – JCPDS file 01-074-1957) indicates the presence of the tetrag-onal phase. Moreover, the intensity ratio I0 0 2/I2 0 0of the two peaks is not equal to 0.5, and shows that the structure of the ceramics consists in a mixture of cubic (or pseudo-cubic) and tetragonal perovskite phases. The magnitude of the peak
split-ting, indicating the tetragonality, decreases as the lanthanum content increases. Since La3+ cations (ionic radius = 1.032 ˚A) substitute for larger Ba2+ cations (ionic radius = 1.35 ˚A)34 on the A sites (in 6 coordinate for the tetragonal phase), there will be a decrease in the lattice value in the c direction, i.e., the c/a ratio tends towards 1.
3.3. Microstructure 3.3.1. Powders
TEM observations of the powder samples (Figs. 4–6) reveal that the grains are mostly equiaxed. Some spherical shaped grains are observed as shown by an arrow inFig. 5a. This shape is observed among the small grains, i.e., for GS less than 27 nm and can be attributed to the cubic phase.22,32Some tetragonal (close to quasi-cubic) and cubic shaped grains (shown by an arrow in Fig. 5b) are also observed in some cases. One or more angles are sometimes rounded (shown by an arrow inFig. 6a). Elongated grains (shown by an arrow inFig. 4) are rarely observed. No defects or dislocations are noticed.
The GS distribution has been determined for each pow-der. The histogram of each sample is shown in Fig. 7. The characteristics of each distribution are reported inTable 2. An asymmetric monomodal dispersion has been observed for each sample, with a minima (Dmin) of 5 nm (x = 0.08 and x = 0.14) and a maxima (Dmax) of 140 nm (pure BTO sample). Only for the samples of composition x = 0.08 and x = 0.14, one can notice a distribution close to symmetric monomodal, i.e., normal dis-tributions. For each GS distribution, the median GS (Dmedian) is equal or very close to the Daverage value. The spreads of overall GS distributions are slightly wide and no significant differences are noticed among them (like it is shown by the values of the relative standard deviation (RSD) of the disper-sions (Table 2) which are approximately of the same order of magnitude).
The variation of Daveragevalue versus La concentration (for the powders) is represented inFig. 8. The Daveragevalue is the same for pure BTO and for x = 0.01 (∼54 nm), but the GS dis-tribution of x = 0.01 sample is a little wider in comparison to the BTO one. The Daverage value decreased slightly (∼42 nm) for x = 0.05 sample but then more significantly for x = 0.08 (∼27 nm). A higher La content (x = 0.14) does not affect the Daveragevalue of the powders.
3.3.2. Ceramics
After sintering at 1050◦C, the ceramic grains remained mostly equiaxed (Figs. 9–11) quite similar to quadratic or cubic forms (with truncated or rounded angles). No particular defect is observed. The grains are monodomain and dislocation free.
The GS distribution has been determined for each ceramic. The histogram of each ceramic sample is shown inFig. 7. The characteristics of each distribution are reported inTable 2. The variation of Daverage value versus La concentration, for these ceramic samples, is represented inFig. 8. Although using SPS technique, and low sintering temperature (1050◦C), little grain growth occurs. The Daveragevalue of the dense ceramics is higher than the one of the initial powders and the Dmaxvalue determined
Fig. 8. Daveragevalue of Ba1−xLaxTiO3−δpowders and ceramics (obtained at
1050◦C) as a function of lanthanum concentration (x). Vertical bars represent σ, the uncertainty of Daverage(σ2=(i=1→N)(Di− Daverage)2/N(N− 1)).
among all compositions is about 600 nm (for x = 0) as shown in Fig. 9.
The GS distributions of x = 0.08 and x = 0.14 samples are quite symmetrical. For the other ceramics, the distributions remain asymmetric. Moreover, the RSD values are smaller than the one of the powders, indicating better homogeneity for the majority of ceramics, except for the composition x = 0.14. The relative increase of the average size (RIAS) after sintering has been deter-mined. It is the most important (RIAS = 511%) for pure BTO. It decreases to 172%, 64% and 33% for the respective compo-sitions of ceramics: x = 0.01, 0.05 and 0.08. The presence of La in BTO seems to slow down the overall diffusion rate during the sintering process. The effect of La on grain growth and thus on the sintering mechanisms (with our sintering conditions), when x was varied from 0.08 to 0.14 can be described as follows: the Daveragevalue and the RIAS for x = 0.08 and x = 0.14 are approx-imately the same. For x = 0.14, the width of the histogram was wider than that for x = 0.08 and the percentage of small grains that were between 7 and 20 nm was higher for x = 0.14 than for x = 0.08. Although prior to sintering, the percentage of grains that ranged from 5 to 20 nm was higher for the powder containing
Fig. 9. Bright field TEM images of Ba1−xLaxTiO3−δceramic samples (obtained at 1050◦C) showing grains with sizes of 600 nm which correspond to the Dmax
Table 2
Grain size (D) of La1−xBaxTiO3−δpowders (Powd.) and ceramics (Ceram.) – the relative increase of average size (RIAS) is also reported. La content (x) 0.00 0.01 0.05 0.08 0.14 Dmin–Dmax(nm) Powd. 20–130 20–130 15–75 6–61 5–60 Ceram. 140–600 43–365 30–110 8–75 8–79 Dmedian(nm) Powd. 54 48 42 26 28 Ceram. 289 143 70 34 38 Daverage(σ)a(nm) Powd. 54 (3) 53 (3) 42 (2) 27 (2) 28 (1) Ceram. 330 (11) 144 (4) 69 (2) 36 (1) 39 (2) Nb Powd. 56 59 57 65 73 Ceram. 82 154 89 101 73 Sc(nm) Powd. 19 22 14 13 10 Ceram. 97 52 15 13 15 RSDd Powd. 0.35 0.41 0.33 0.48 0.36 Ceram. 0.29 0.36 0.19 0.36 0.38 RIASe(%) 511 172 64 33 39
aσ: uncertainty of the average size (σ2=
(i=1→N)(Di− Daverage)2/N(N− 1)). b N: number of measurements.
cS: standard deviation of the dispersion (S2=
(i=1→N)(Di− Daverage)2/(N− 1)). d RSD: relative standard deviation of the dispersion (RSD = S/D
average). eRIAS: relative increase of average size = [D
average (ceram.)− Daverage (powd.)]/Daverage (powd.).
0.08 La than for powder containing 0.14 La which could mean that the grain growth was slower for x = 0.14 than for x = 0.08. In addition to that, the densification curves (Fig. 1) showed that the densification rate is higher for the sample containing the smaller amount of La, i.e., x = 0.08 and the density of sample containing 0.08 La became stable after 1 min of dwell time but for the other sample (i.e., x = 0.14), the density continued to increase along the totality of the dwell time. So, the increase of La content from x = 0.08 to x = 0.14 had an effect both on the global GS and on the densification rate.
4. Discussion
The mechanisms occurring during the powder sinter-ing are complex and depend on many factors (heatsinter-ing rate, applied pressure, particle size and distribution, etc.). All the processes that are involved during SPS sinter-ing are not fully understood and some mechanisms are still under debate, such as plasma generation.40–42 Among the mechanisms that are generally assumed during SPS sintering, we can mention: the surface, grain boundary
Fig. 10. Bright field TEM images of Ba1−xLaxTiO3−δceramic samples (sintered at 1050◦C): (a) for x = 0 showing grains which are equiaxed with a Daveragevalue of
330 nm; (b) and (c) for x = 0.01 showing equiaxed polydispersed grains which are smaller (Daverage= 144 nm) than that for x = 0.00 and are quite similar to tetragonal
Fig. 11. Bright field TEM images of Ba1−xLaxTiO3−δceramic samples (sintered at 1050◦C): (a) for x = 0.05 showing grains which are equiaxed with a Daverage
value of 69 nm; (b) for x = 0.08 showing equiaxed grains (Daverage= 36 nm).
and volume diffusion; vaporization–condensation and plastic deformation.
In all our samples, the structures of initial powders are cubic (or pseudo-cubic) which is common for this type of nanograins BTO perovskites.28,31–33We have also observed an increase of the lattice parameter (displacing X-ray peak positions towards lower 2θ values) when x varies from 0.05 to 0.08, i.e., when the Daveragevalue was decreased from 42 nm to 27 nm as described by several authors.35–39The Daveragevalue of the starting pow-ders lies between 54 and 27 nm. FromFig. 1, one can observe that the densification starts at lower temperature as the GS decreases. For example, the sintering onset appears at 762◦C for x = 0.14 (Daverage= 28 nm) and 779◦C for x = 0.08 (Daverage= 27 nm). This behavior could be related to the high surface energy and a surface diffusion in the early stages of sintering. At this stage,
particle growth is limited by the high heating rate, thus favorizing the surface diffusion of species. As the GSs of green samples are small, the surface diffusion is preponderant and the densifica-tion starts early.43So, the sintering onset temperature increases systematically with the increase of the initial GS of samples, i.e., with the decrease of La content (805◦C for x = 0.05, 812◦C for x = 0.01) and reaches 855◦C for the sample showing the largest initial particle size, i.e., pure BTO (Daverage= 54 nm). As the shrinkage begins and the temperature is further increased, the sintering is not only limited by surface diffusion anymore, but also limited by the bulk diffusion rate. Lanthanum doping obviously influences the densification rate, thus the sintering behavior. It is well known that BTO doped with a small amount of donor cations as lanthanum (>0.5 mol%), yields to cation vacancy compensation, i.e., oxygen vacancy and results in the material being fine-grained. When the lanthanum percentage increases, the number of oxygen vacancies becomes important and contributes to decrease the overall cationic transport, thus diminishing the diffusion rate and in consequence the densifica-tion rate.44,45
For the chosen sintering conditions, undoped BTO reached a maximum densification of 97% with a RIAS of grains of about 511%. For low La content (x = 0.01), with similar initial GS, the densification does not change, but the RIAS of grains is only 172%. This difference in the value of RIAS, clearly indi-cates that a small amount of La-doping has a big influence on the grain growth during the SPS sintering of Ba1−xLaxTiO3−δ. This behavior could be the result of the formation of the oxy-gen vacancies by the introduction of La and also, the oxyoxy-gen vacancies that are probably created during sintering, due to the reducing atmosphere inherent to the SPS process (vacuum in presence of the graphite die). The grain growth decreases as the La content increases (RIAS = 64% for x = 0.05 and 33% for x = 0.08).
The ceramics of compositions x = 0.08 and x = 0.14 show quite a different behavior, the densification is very low, only 85%. It is important to note that the RIASs of those ceramics are the lowest of the series, respectively 33% and 39%. As a conse-quence, the values of their Daverageare very low, 36 and 39 nm. This result is in agreement with previously reported results,41,43 indicating that densification is enhanced by grain growth.
The variation of the Daverage value versus La concentra-tion (for ceramics) is plotted in Fig. 8: for 0 < x < 0.08 the Daveragevalue of ceramics was highly decreased (in this zone, the graph was similar to an exponential decay curve) and for 0.08 < x < 0.14, the Daveragevalue of ceramics was approximately constant (about 36 nm). Then, these results showed that when the amount of La was increased from x = 0 to x = 0.08, the RIAS was decreased with the enhancement of La content, but, from x = 0.08 to x = 0.14, the increasing of La-doping did not show any significant effect on Daveragevalue. But, the observation of the overall results (histograms and densification curves) of x = 0.08 and x = 0.14 samples had shown that the grain growth in x = 0.14 sample occurred more slowly than that in x = 0.08 sample (the percentage of small grains ranged from 7 to 20 nm, observed in x = 0.14 sample was higher than those observed in x = 0.08 sample). In addition to that, the densification curves (Fig. 1) had
shown a decrease of densification rate when the composition was changed from x = 0.08 to x = 0.14.
Two hypotheses can be suggested in order to explain why the Daveragevalue of the ceramics did not change when x was increased from 0.08 to 0.14:
- The first hypothesis which is the more probable, is based on the fact that from a critical value of x (i.e., x = 0.08), the RIAS of grains (determined after sintering) reached its pos-sible minimal value (with our sintering conditions) in a way that even if the La content was increased, the RIAS of grains cannot decrease any more. In other words, we can see that for x = 0.08, the initial GSs were already fine (between 6 and 61 nm with Daverage= 27 nm) and their GSs were increased only very slightly after sintering (between 8 and 75 nm with Daverage= 36 nm). So, there is an increase in the Daveragevalue of only 9 nm, which is very small. It is thus highly probable that if we increase the La content, the RIAS of grains after sintering will not change because it has reached a limit value. - The second hypothesis is to consider that a change in La-doping mechanism may occur for a critical value of x (i.e., x = 0.08). It is probable that for x < 0.08, La3+ could occupy Ba-sites and for x≥ 0.08, La3+could occupy both Ba-sites and Ti-sites.46,47However, this hypothesis is less probable than the previous one because lanthanum, among the rare earth elements, tends to occupy the Ba-sites.26–28,47–49
5. Conclusion
Ba1−xLaxTiO3−δ ceramics with (0.00≤ x ≤ 0.14) and with Daverage value ranging from 330 (11) to 36 (1) nm were obtained using the SPS technique (Tsintering= 1050◦C, dwell time = 3 min) to nanometric powders (prepared by coprecipita-tion followed by a calcinacoprecipita-tion at 850◦C) with Daverage values ranging from 54 (3) to 27 (2) nm. Both powders and ceram-ics were characterized by XRD and TEM. The initial powders correspond to the cubic (or pseudo-cubic) perovskite phase. An increase of the lattice parameter was observed when x varied from 0.05 to 0.08 (i.e., when the Daveragevalue was decreased from 42 nm to 27 nm). The structure of ceramics consists in a mixture of cubic (or pseudo-cubic) and tetragonal perovskite type phases (the tetragonality decreases with the enhancement of x). Lanthanum doping influences the densification rate, i.e., the sintering behavior. This behavior could be the result of the formation of the oxygen vacancies by the introduction of La. The highest is the value of x, the lowest are the temperature of sintering onset, densification rate, grain growth and thus density of ceramics. Doping in the 0 < x < 0.08 range diminishes con-siderably the Daveragevalue of ceramics (from 330 nm to 36 nm) and for 0.08 < x < 0.14, the Daveragevalue of ceramics remains approximately constant. Two hypotheses are proposed to explain why the Daveragevalue of ceramics did not change when x was increased from 0.08 to 0.14: either from a critical value of x (i.e., x = 0.08), the lowest limit value of the RIAS of grains could be attained (with our sintering conditions), either a change in La-doping mechanism might occur for x≥ 0.08 (for x < 0.08, La3+ could occupy Ba-sites and for x≥ 0.08, La3+could occupy both
Ba-sites and Ti-sites). But the second hypothesis is less probable than the first one because that among the rare earth elements, lanthanum, in most cases, tends to occupy Ba-sites.
Acknowledgment
The authors would like to thank Mr. Laurent Weingarten (Service Commun de Microscopie Electronique TEMSCAN – Université Paul Sabatier – Toulouse) for his kind help in the preparation of samples for TEM observations.
References
1. Kishi H, Mizuno Y, Chazono H. Base-metal electrode-multilayer ceramic capacitors: past, present and future perspectives. Jpn J Appl Phys 2003;42(Part 1, No. 1):1–15.
2. Haertling GH. Ferroelectric ceramics: history and technology. J Am Ceram Soc 1999;82(4):797–818.
3. Kutty TRN, Hari NS. Multifunctional cryogenic sensors from n-BaTiO3
ceramics having strong negative temperature coefficient of resistance. J Phys D: Appl Phys 1995;28(2):371–4.
4. Lin PT, Yi F, Ho ST, Wessels BW. Two-dimensional ferroelectric photonic crystal waveguides: simulation, fabrication, and optical characterization. J Lightwave Technol 2009;27(19):4330–7.
5. Luan W, Gao L, Kawaoka H, Sekino T, Niihara K. Fabrication and charac-teristics of fine-grained BaTiO3ceramics by spark plasma sintering. Ceram
Int 2004;30(3):405–10.
6. Takeuchi T, Bétourné E, Tabuchi M, Kageyama H, Kobayashi Y, Coats A, et al. Dielectric properties of spark-plasma-sintered BaTiO3. J Mater Sci
1999;34:917–24.
7. Takeuchi T, Tabuchi M, Kageyama H, Suyama Y. Preparation of dense BaTiO3ceramics with submicrometer grains by spark plasma sintering. J
Am Ceram Soc 1999;82(4):939–43.
8. Takeuchi T, Suyama Y, Sinclair DC, Kageyama H. Spark-plasma-sintering of fine BaTiO3 powder prepared by a sol–crystal method. J Mater Sci
2001;36:2329–34.
9. Takeuchi T, Capiglia C, Balakrishnan N, Takeda Y, Kageyama H. Prepara-tion of fine-grained BaTiO3ceramics by spark plasma sintering. J Mater
Res 2002;17(3):575–81.
10. Wang X, Deng X, Li L. Processing, microstructure and properties of nanograin barium titanate ceramics by spark plasma sintering. Mater Sci Forum 2009;606(Advances in Ceramic Materials):135–47.
11. Zhao Z, Buscaglia V, Bowen P, Nygren M. Spark plasma sintering of nano-crystalline ceramics. Key Eng Mater 2004;264–268(Pt. 3, Euro Ceramics VIII):2297–300.
12. Buscaglia MT, Buscaglia V, Viviani M, Petzelt J, Savinov M, Mitoseriu L, et al. Ferroelectric properties of dense nanocrystalline BaTiO3ceramics.
Nanotechnology 2004;15(9):1113–7.
13. Li B, Wang X, Li L, Zhou H, Liu X, Han X, et al. Dielectric properties of fine-grained BaTiO3prepared by spark-plasma sintering. Mater Chem
Phys 2004;83(1):23–8.
14. Maiwa H. Preparation and properties of BaTiO3ceramics by spark plasma
sintering. Jpn J Appl Phys 2008;47(9, Pt. 2):7646–9.
15. Li B, Wang X, Cai M, Hao L, Li L. Densification of uniformly small-grained BaTiO3 using spark-plasma-sintering. Mater Chem Phys
2003;82(1):173–80.
16. Lu DY, Sun XY, Toda M. A novel high-k ‘Y5V’ barium titanate ceramics co-doped with lanthanum and cerium. J Phys Chem Solids 2007;68(4):650–64. 17. Morrison FD, Sinclair DC, West AR. Electrical and structural char-acteristics of lanthanum-doped barium titanate ceramics. J Appl Phys 1999;86(11):6355–66.
18. Morrison FD, Sinclair DC, Skakle JMS, West AR. Novel doping mecha-nism for very-high-permittivity barium titanate ceramics. J Am Ceram Soc 1998;81(7):1957–60.
19. Liu Y, Feng Y, Wu X, Han X. Microwave absorption properties of La doped barium titanate in X-band. J Alloys Compd 2009;472(1–2):441–5. 20. Jung DS, Koo HY, Jang HC, Kang YC. Effects of La content on the
prop-erties of Ba1−xLaxTiO3powders prepared by spray pyrolysis. Met Mater
Int 2009;15(5):809–14.
21. Jung DS, Lee SH, Kang YC. Effects of dopants on grain growth of nano-sized BaTiO3powders prepared by citric acid-assisted spray pyrolysis. J
Ceram Proc Res 2008;9:307–10.
22. Xinle Z, Zhimei M, Zuojiang X, Guang C. Preparation and characterization on nano-sized barium titanate powder doped with lanthanum by sol–gel process. J Rare Earths 2006;24(Suppl. 1):82–5.
23. Morrison FD, Sinclair DC, West AR. Doping mechanisms and elec-trical properties of La-doped BaTiO3 ceramics. Int J Inorg Mater
2001;3(8):1205–10.
24. Wenhu Y, Yongping PU, Xiaolong C, Jinfei W. Study of reoxida-tion in heavily La-doped barium titanate ceramics. J Phys Conf Ser 2009;152(1):012040.
25. Kuwabara M, Matsuda H, Kurata N, Matsuyama E. Shift of the Curie point of barium titanate ceramics with sintering temperature. J Am Ceram Soc 1997;80(10):2590–6.
26. Kishi H, Kohzu N, Mizuno Y, Iguchi Y, Sugino J, Ohsato H, et al. Effect of occupational sites of rare-earth elements on the microstructure in BaTiO3.
Jpn J Appl Phys 1999;38(9B):5452–6.
27. Xue LA, Chen Y, Brook RJ. The influence of ionic radii on the incorpo-ration of trivalent dopants into barium titanate (BaTiO3). Mater Sci Eng B
1988;B1(2):193–201.
28. Buscaglia MT, Buscaglia V, Viviani M, Nanni P, Hanuskova M. Influ-ence of foreign ions on the crystal structure of BaTiO3. J Am Ceram Soc
2000;20(12):1997–2007.
29. Valdez-Nava Z, Guillemet-Fritsch S, Tenailleau C, Lebey T, Durand B, Chane-Ching JY. Colossal dielectric permittivity of BaTiO3-based
nanocrystalline ceramics sintered by spark plasma sintering. J Electro-ceram 2009;22(1–3):238–44.
30. Abramoff MD, Magelhaes PJ, Ram SJ. Image processing with ImageJ. Biophotonics Int 2004;11(7):36–42.
31. Vijayalakshmi R, Rajendran V. Synthesis and characterization of cubic BaTiO3nanorods via facile hydrothermal method and their optical
proper-ties. Dig J Nanomater Bios 2010;5:511–7.
32. Zhu XH, Zhu JM, Zhou SH, Liu ZG, Ming NB, Hesse D. Microstruc-tural characterization of BaTiO3ceramic nanoparticles synthesized by the
hydrothermal technique. Diffus Defect Data B 2005;106(From Nanopow-ders to Functional Materials):41–6.
33. McHale JM, McIntyre PC, Sickafus KE, Coppa NV. Nanocrys-talline BaTiO3 from freeze-dried nitrate solutions. J Mater Res
1996;11(5):1199–209.
34. Shannon RD. Revised effective ionic radii and systematic studies of interatomic distances in halides and chalcogenides. Acta Crystallogr A 1976;A32(5):751–67.
35. Huang TC, Wang MT, Sheu HS, Hsieh WF. Size-dependent lattice dynamics of barium titanate nanoparticles. J Phys Condens Matter 2007;19(47):476212.
36. Valot C, Floquet N, Mesnier M, Niepce JC. Diffraction des rayons X par un polycristallin et microstructure en domaines ferroélectriques. J Phys IV France 1996;06(C4):71–89.
37. Perriat P. The Laplace law in finely divided powders. Nanostructured Mater 1995;6(5–8):791–4.
38. Caboche G, Niepce JC. Dielectric constant of fine grain BaTiO3. In: Nair
KM, Guha JP, Okamoto A, editors. Ceramics transactions 32: dielec-tric ceramics: processing, properties and applications. Westerville, OH: American Ceramic Society; 1993. p. 339–45.
39. Wada S, Yasuno H, Hoshina T, Nam SM, Kakemoto H, Tsurumi T. Prepa-ration of nm-sized barium titanate fine particles and their powder dielectric properties. Jpn J Appl Phys 2003;42(9B):6188–95.
40. Chaim R. Densification mechanisms in spark plasma sintering of nanocrys-talline ceramics. Mater Sci Eng A 2007;443(1–2):25–32.
41. Kim HT, Han YH. Sintering of nanocrystalline BaTiO3. Ceram Int
2004;30(7):1719–23.
42. Aman Y, Garnier V, Djurado E. A screening design approach for the understanding of spark plasma sintering parameters: a case of translucent polycrystalline undoped alumina. Int J Appl Ceram Tech 2010;7(5):574– 86.
43. Haussonne JM, Carry C, Bowen P, Barton J. Ceramics and glasses. In: Principles and techniques of manufacture. PPUR; 2005.
44. Shirasaki S, Yamamura H, Haneda H, Kakegawa K, Moori J. Defect struc-ture and oxygen diffusion in undoped and La-doped polycrystalline barium titanate. J Chem Phys 1980;73(9):4640–5.
45. Shimanskij AF, Drofenik M, Kolar D. Subsolidus grain growth in donor doped barium titanate. J Mater Sci 1994;29(23):6301–5.
46. Reddy BS, Rao PK, Rao RMS. Effect of La substitution on the struc-tural and dielectric properties of BaZr0.1Ti0.9O3ceramics. J Alloys Compd
2009;481(1–2):692–6.
47. Watanabe K, Ohsato H, Kishi H, Okino Y, Kohzu N, Iguchi Y, et al. Solubility of La–Mg and La–Al in BaTiO3. Solid State Ionics
1998;108(1–4):129–35.
48. Itoh JI, Park DC, Ohashi N, Sakaguchi I, Yashima I, Haneda H, et al. Oxygen defects related to electrical properties of La-doped BaTiO3. Jpn J
Appl Phys 2002;41(6A):3798–803.
49. Tsur Y, Hitomi A, Scrymgeour I, Randall CA. Site occupancy of rare-earth cations in BaTiO3. Jpn J Appl Phys 2001;40(1):255–8.
